feat: 16-bit TIFF viewer with composable enhancement pipeline

[Copilot]

Implements a full-stack tool for viewing and interactively processing 16-bit grayscale TIFF images with a composable, reorderable enhancement pipeline.

Backend — FastAPI + Python (backend/)

• POST /api/upload — reads 16-bit TIF via tifffile, stores as float32 ndarray, returns file_id
• POST /api/process — applies ordered enhancement pipeline, returns base64 PNG + 256-bin histogram
• GET /api/image/{file_id} / GET /api/histogram/{file_id} — individual render/histogram endpoints
• processors/histogram_eq.py — skimage.equalize_hist on float-normalised 16-bit data
• processors/local_contrast_norm.py — (x − μ_local) / (σ_local + ε) via Gaussian blur, rescaled to [0, 65535]

Frontend — Vue 3 + Vite (frontend/)

• EnhancementPanel — drag-to-reorder pipeline (vuedraggable), add from dropdown, remove with ×
• ParameterPanel / SliderRow — LCN exposes Sigma (0.30–32), Epsilon (log-scale, 1e-4–0.5), Output Gain (0–4); changes debounced 300 ms before re-processing
• Histogram — Chart.js bar over full 16-bit range; tooltip: Value X / Distribution Y
• ImageViewer — scroll-to-zoom, drag-to-pan, fit-to-screen
• Toolbar — file open, zoom controls, annotation mode placeholders
• Vite dev proxy: /api → http://localhost:8000